Kirsch算子地质图像边缘检测算法并行化研究  

Research on Parallelization of Kirsch Operator Edge Detection Algorithm for Geological Image

作　　者：田宸玮 王雪纯 杨嘉能 钱育蓉[1,2,3] TIAN Chenwei;WANG Xuechun;YANG Jianeng;QIAN Yurong(School of Software,Xinjiang University,Urumqi Xinjiang 830008,China;Key Laboratory of Signal Detection and Processing in Xinjiang Uygur Autonomous Region,Urumqi Xinjiang 830008,China;Key Laboratory of Software Engineering,Xinjiang University,Urumqi Xinjiang 830008,China;Research Institute of Exploration and Development,Tuha Oil Field Branch Company Ltd.of Petro China,Hami Xinjiang 839009,China)

机构地区：[1]新疆大学软件学院,新疆乌鲁木齐830008 [2]新疆维吾尔自治区信号检测与处理重点实验室,新疆乌鲁木齐830008 [3]新疆大学软件工程重点实验室,新疆乌鲁木齐830008 [4]中国石油吐哈油田公司勘探开发研究院,新疆哈密839009

出　　处：《新疆大学学报（自然科学版）（中英文）》2021年第1期54-60,68,共8页Journal of Xinjiang University(Natural Science Edition in Chinese and English)

基　　金：国家自然科学基金(61966035);国家自然科学基金联合基金-重点项目(U1803261);新疆维吾尔自治区科技厅国际合作项目(2020E01023)。

摘　　要：针对大尺寸地质图像边缘检测算法计算密集和数据密集的特性,为提高地质图像边缘检测算法的计算效率,提出一种自适应阈值的Kirsch算子的边缘检测算法.从传统算法层面,通过减少运算次数以及针对阈值设定随机性较大的问题提出自适应阈值的方法对其进行优化.从算法并行层面,在CPU-GPU传输开销以及线程规模选取上分析优化.经测试,改进的算法比现有算法减少了计算量,获取的边缘更清晰,对大于2 048×2 048尺寸的地质图像加速比可以保持在80倍以上(不考虑传输开销可保持在300倍以上).该方法的并行较易实现,为实时在线的地质图像边缘检测提供了可能.Aiming at the computation-intensive and data-intensive characteristics of the edge detection algorithm of large-scale geological images,in order to improve the computational efficiency of the edge detection algorithm of geological images,an adaptive threshold kirsch operator edge detection algorithm was proposed.From the perspective of traditional algorithm,an adaptive threshold method is proposed to optimize the algorithm by reducing the number of operations and aiming at the problem of setting the random threshold.At the parallel level,cpu-gpu transmission overhead and thread size were selected for analysis and optimization.Compared with the existing algorithm,the improved algorithm requires less computation and gets clearer edges.The acceleration ratio of the geological image larger than 2048×2048 can be maintained at more than 80 times(the transmission cost can be maintained at more than 300 times).The parallel optimization scheme is easy to implement and can be applied to the edge detection of online real-time geological images.

关 键 词：边缘检测 KIRSCH算子 并行计算 地质图像 统一计算设备架构 

分 类 号：TP391.413[自动化与计算机技术—计算机应用技术] TP751.1[自动化与计算机技术—计算机科学与技术]